Limitations on causal mapping in evaluation - data quality

For example, it might be agreed that a reasonable threshold of evidence that C influenced E is that (i) a specified number or proportion of respondents independently mentioned the link; (ii) the connection can be plausibly explained theoretically and (iii) researchers anticipated confirmation bias and other potential sources of bias and took adequate steps to mitigate against them. See Tegarden et al. (2016) for a discussion of the role of anonymity in causal mapping in organisations.